From: Peter Zijlstra <a.p.zijlstra@chello.nl>
To: Eric Dumazet <dada1@cosmosbay.com>
Cc: "David S. Miller" <davem@davemloft.net>,
Andrew Morton <akpm@linux-foundation.org>,
linux kernel <linux-kernel@vger.kernel.org>,
netdev@vger.kernel.org, Christoph Lameter <clameter@sgi.com>,
"Zhang, Yanmin" <yanmin_zhang@linux.intel.com>
Subject: Re: [PATCH] alloc_percpu() fails to allocate percpu data
Date: Thu, 21 Feb 2008 23:26:05 +0100 [thread overview]
Message-ID: <1203632765.6112.20.camel@lappy> (raw)
In-Reply-To: <47BDBC23.10605@cosmosbay.com>
On Thu, 2008-02-21 at 19:00 +0100, Eric Dumazet wrote:
> Some oprofile results obtained while using tbench on a 2x2 cpu machine
> were very surprising.
>
> For example, loopback_xmit() function was using high number of cpu
> cycles to perform the statistic updates, supposed to be real cheap
> since they use percpu data
>
> pcpu_lstats = netdev_priv(dev);
> lb_stats = per_cpu_ptr(pcpu_lstats, smp_processor_id());
> lb_stats->packets++; /* HERE : serious contention */
> lb_stats->bytes += skb->len;
>
>
> struct pcpu_lstats is a small structure containing two longs. It
> appears that on my 32bits platform, alloc_percpu(8) allocates a single
> cache line, instead of giving to each cpu a separate cache line.
>
> Using the following patch gave me impressive boost in various
> benchmarks ( 6 % in tbench) (all percpu_counters hit this bug too)
>
> Long term fix (ie >= 2.6.26) would be to let each CPU allocate their
> own block of memory, so that we dont need to roudup sizes to
> L1_CACHE_BYTES, or merging the SGI stuff of course...
>
> Note : SLUB vs SLAB is important here to *show* the improvement, since
> they dont have the same minimum allocation sizes (8 bytes vs 32
> bytes). This could very well explain regressions some guys reported
> when they switched to SLUB.
I've complained about this false sharing as well, so until we get the
new and improved percpu allocators,
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
> Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
>
> mm/allocpercpu.c | 15 ++++++++++++++-
> 1 files changed, 14 insertions(+), 1 deletion(-)
>
>
> plain text document attachment (percpu_populate.patch)
> diff --git a/mm/allocpercpu.c b/mm/allocpercpu.c
> index 7e58322..b0012e2 100644
> --- a/mm/allocpercpu.c
> +++ b/mm/allocpercpu.c
> @@ -6,6 +6,10 @@
> #include <linux/mm.h>
> #include <linux/module.h>
>
> +#ifndef cache_line_size
> +#define cache_line_size() L1_CACHE_BYTES
> +#endif
> +
> /**
> * percpu_depopulate - depopulate per-cpu data for given cpu
> * @__pdata: per-cpu data to depopulate
> @@ -52,6 +56,11 @@ void *percpu_populate(void *__pdata, size_t size, gfp_t gfp, int cpu)
> struct percpu_data *pdata = __percpu_disguise(__pdata);
> int node = cpu_to_node(cpu);
>
> + /*
> + * We should make sure each CPU gets private memory.
> + */
> + size = roundup(size, cache_line_size());
> +
> BUG_ON(pdata->ptrs[cpu]);
> if (node_online(node))
> pdata->ptrs[cpu] = kmalloc_node(size, gfp|__GFP_ZERO, node);
> @@ -98,7 +107,11 @@ EXPORT_SYMBOL_GPL(__percpu_populate_mask);
> */
> void *__percpu_alloc_mask(size_t size, gfp_t gfp, cpumask_t *mask)
> {
> - void *pdata = kzalloc(nr_cpu_ids * sizeof(void *), gfp);
> + /*
> + * We allocate whole cache lines to avoid false sharing
> + */
> + size_t sz = roundup(nr_cpu_ids * sizeof(void *), cache_line_size());
> + void *pdata = kzalloc(sz, gfp);
> void *__pdata = __percpu_disguise(pdata);
>
> if (unlikely(!pdata))
next prev parent reply other threads:[~2008-02-21 22:26 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-02-21 18:00 [PATCH] alloc_percpu() fails to allocate percpu data Eric Dumazet
2008-02-21 22:26 ` Peter Zijlstra [this message]
2008-02-23 9:23 ` Nick Piggin
2008-02-27 19:44 ` Christoph Lameter
2008-03-03 3:14 ` Nick Piggin
2008-03-03 7:48 ` Eric Dumazet
2008-03-03 9:41 ` Nick Piggin
2008-03-03 19:30 ` Christoph Lameter
2008-02-23 8:04 ` Andrew Morton
2008-02-27 19:59 ` Christoph Lameter
2008-02-27 20:24 ` Andrew Morton
2008-02-27 21:56 ` Christoph Lameter
2008-03-01 13:53 ` Eric Dumazet
2008-03-11 18:15 ` Mike Snitzer
2008-03-11 18:41 ` Eric Dumazet
2008-03-11 19:39 ` Mike Snitzer
2008-03-12 0:18 ` [stable] " Chris Wright
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1203632765.6112.20.camel@lappy \
--to=a.p.zijlstra@chello.nl \
--cc=akpm@linux-foundation.org \
--cc=clameter@sgi.com \
--cc=dada1@cosmosbay.com \
--cc=davem@davemloft.net \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=yanmin_zhang@linux.intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).